This message may have been sent using an unsubscribed account. My apologies.
----
I would like guidance as to what the rules are regarding changing, overwriting, and generation of RPM installed conf files within the system-config tools?
Specifically, should the original config files remain in tact aside from changes within the tool? Can they be moved to a backup file, and a new one created? What are the guidelines for this?
Obviously all config files should be parsed, and all current directives worked into the presentation of existing configurations. Should whitespace, comments and erroneous hand-edited lines and the like also be preserved?
I've created some python configuration import scripts for: /var/yp/Makefile /etc/ypserv.conf and other related files. Parsing and tokenizing of directives is done, as well as directive validation. I would like to know what of the other "stuff" in the config files needs to be retained.
I have another thing that I would like guidance on as well. In developing system-config-nis, I will want to give administrators the options to publish autofs, services, aliases (and other NIS map types) through the GTK and Newt interfaces. These file do not belong (rpm -qf) to ypserv, or the other yp* packages. What is the policy regarding touching files which belong to other packages? As an example, /etc/services belongs to the setup package, but is very relevant to NIS configuration. Is it reasonable to adjust those files which will be used to create maps within the NIS config tools, or would that be stepping over the proverbial line?
Thanks,
Mike